The present invention relates to modified enzyme acceptor polypeptide fragments of .beta.-galactosidase which have increased activity, to processes for the preparation thereof, and to the use thereof as reagents in enzyme complementation immunoassays.
A number of homogeneous immunoassays have recently been described that utilize the complementation or reassociation of enzymatically-inactive polypeptide fragments to form active enzymes as a step of generating a detectable signal which can be utilized to determine the amount of an analyte of interest that may be present in a sample such as blood serum. Several of these assays propose utilizing the enzyme .beta.-galactosidase as the enzyme formed by complementation.
Enzyme complementation involves the association of two or more inactive polypeptides which together provide the structural information required for the formation of a biologically active enzyme complex resembling that of the native parent enzyme. The enzymatically-inactive polypeptide fragments can be obtained as the result of proteolysis, chemical cleavage, chemical synthesis, or as the result of a missense or nonsense mutation of the gene coding for the active enzyme. Examples of protein complementation systems which yield an enzymatically-active complex are the ribonuclease-S' complex, the staphylococcal nuclease T complex, various two- and three-fragment complexes derived from cytochrome c, and the alpha- and omega-complementation complexes of E. coli .beta.-galactosidase. The interactions which stabilize these complexes are non-covalent in nature and are similar to those involved in the formation and maintenance of the three-dimensional structure of the native enzyme.
Enzyme complementation has been utilized as the underlying basis for the development of a novel homogeneous immunoassay technology. Farina and Golke, U.S. Pat. No. 4,378,428 issued Mar. 29, 1983, and Gonelli et al., (1981, Biochem. and Biophys. Res. Commun. 102:917-923) describe an immunoassay based upon the reassociation of S-peptide and S-protein, both of which are derived from the proteolytic cleavage of ribonuclease A, to generate ribonuclease catalytic activity. Specific components of the assay system include an analyte covalently attached to the S-peptide (amino acids 1-20), free S-protein (amino acids 21-124), an antibody specific for the analyte, and a substrate of ribonuclease which is capable of being converted to a reporter molecule. The anti-analyte antibody inhibits the association of the analyte:S-peptide conjugate with the S-protein, thereby reducing the level of enzymatically-active complex and thus the signal generated by the enzymatic reaction. In the presence of a sample containing free analyte, a competition for the antigen binding site occurs between sample-born analyte and the S-peptide conjugate. The concentration of S-peptide conjugate free to participate in complementation with the S-protein fragment, and the resulting signal due to the enzymatic activity of the ribonuclease A' complex, are directly proportional to the concentration of free analyte in the sample.
A similar immunoassay system based on the alpha-complementation system of E. coli .beta.-galactosidase polypeptide fragments is described in Henderson, U.S. Pat. No. 4,708,929, issued Nov. 24, 1987, and Henderson, PCT Appl. No. PCT/US90/02491, published Nov. 15, 1990, both of which are incorporated herein by reference. .beta.-Galactosidase alpha-complementation involves the association of an alpha-acceptor polypeptide fragment and an alpha-donor polypeptide fragment and the subsequent formation of an enzymatically active .beta.-galactosidase molecule. The alpha-acceptor is derived from the internal deletion or chain interruption of consecutive amino acids located within the N-terminus proximal segment of the .beta.-galactosidase molecule. Specific examples include the lac Z M15 .beta.-galactosidase deletion mutant lacking residues 11-41 of the wild-type sequence, and the lac Z M112 mutant in which residues 23-31 have been deleted. The alpha-donor polypeptide can be derived from chemical or proteolytic cleavage of the wild-type protein. The cyanogen bromide fragment CNBr2 composed of amino acid residues 3-92, or the V8 protease peptide spanning residues 3-40, both possess alpha-donor activity.
Alpha-donor and alpha-acceptor polypeptides can also be generated through the application of recombinant DNA technology and peptide synthesis techniques. A readily available supply of these molecules and the ability to modify the structure of either the alpha-donor or the alpha-acceptor polypeptides through these techniques has led to the development of an optimized complementation system which has been employed in cloned enzyme donor-based homogeneous immunoassays. The alpha-donor molecule can be chemically coupled with a specific analyte of interest through the modification of either a cysteine or lysine residue which has been suitably located within the sequence of the alpha-donor molecule such that the conjugation does not interfere with the complementation reaction. Complementation between the alpha-acceptor and alpha-donor can be modulated by an antigen-antibody reaction between an analyte-specific antibody and the alpha-donor to which an analyte has been conjugated. In the presence of free analyte, a competition between the free and alpha-donor-conjugated analyte is established for the antigen binding site of the antibody. Thus, an increase in the level of free analyte results in an elevation in the quantity of alpha-donor conjugate which is available for complementation with alpha-acceptor. As a result, the concentration of the alpha-acceptor:alpha-donor complex and reporter molecule produced from the reconstituted enzymatic activity increase and are proportional to the concentration of the free analyte present in the sample. A dose response curve can be constructed by following the activity, i.e., the slope of the rate of the reaction, at several different concentrations of free analyte. The enzyme activity observed at an infinite concentration of free analyte or in the absence of antibody is defined as the "open rate" and represents the maximal signal obtainable from the assay system.
Krevolin and Kates, European Appl. No. 92304354.1, published Nov. 19, 1992, the content of which is herein incorporated by reference, describe enzyme complementation assays involving complementation in the omega region of .beta.-galactosidase between two polypeptide fragments of the whole .beta.-galactosidase molecule formed by a break in the primary structure of .beta.-galactosidase in the omega region. As in alpha complementation, in some cases the two fragments are not strictly complementary so as to form an exact .beta.-galactosidase amino acid sequence without gaps or overlaps; both gaps and overlaps are possible as long as the resulting fragments can assemble into an active .beta.-galactosidase molecule. Like the alpha-acceptor, the omega-acceptor polypeptide is the larger of the two fragments and normally contains about two-thirds of the amino acid sequence of the natural or modified, full-length .beta.-galactosidase. The omega-donor molecule is the smaller fragment containing the remaining one-third (approximately) of the amino acid sequence; the omega-donor molecule is derived from the C-terminus of the .beta.-galactosidase molecule.
.beta.-Galactosidase is a tetrameric protein having a molecular weight of about 540,000 daltons. The four identical monomers consist of 1023 amino acids, each with a molecular weight of 116,000 daltons. The monomeric protein is divided into three regions: the N-terminus proximal segment (the alpha region), a middle region, and a C-terminus distal segment (the omega region).
E. coli .beta.-galactosidase is derived from the Z gene of the lac operon and catalyzes the hydrolysis of .beta.-D-galactopyranosides. The catalytic mechanism of this enzyme involves the general acid catalysis of the glycosidic ester linkage of a substrate molecule by tyrosine-503. This is followed by the loss of the aglycon moiety and the stabilization of a putative carbonium ion intermediate through an interaction with glutamate-461. The final step in the catalytic cycle involves the transgalactosylation of an acceptor molecule, usually water, and the removal of the product from the active site. The active enzyme is composed of four identical subunits with one active site per subunit. Monovalent cations, although not required for activity, dramatically enhance the rate of enzyme catalysis, whereas divalent cations, e.g., Mg.sup.2+ or Mn.sup.2+, are required for activity.
The E. coli .beta.-galactosidase homotetramer contains 64 cysteine residues (16 cysteine residues per subunit), none of which are involved in either the enzymatic activity or the maintenance of the quaternary structure through intersubunit disulfide bridges, as indicated by the stabilization of the molecule in high concentrations of reducing agents. The efficiency of the in vitro association of individual monomers to form the active tetramer is dramatically increased under conditions in which the cysteines are fully reduced. Similarly, reducing agents greatly enhance enzyme complementation. The alpha-acceptor polypeptide contains all 16 cysteine residues present in a single .beta.-galactosidase subunit. However, alpha-acceptor molecules exist as homodimers in solution. Thus, the surface area normally buried at the dimer-dimer interface in .beta.-galactosidase is exposed in the alpha-acceptor. Chemical modification studies of .beta.-galactosidase with iodoacetate lead to the identification of cysteine-500 and cysteine-1021 as surface accessible residues in .beta.-galactosidase (Jornvall et al., 1978, Biochem. 17, 5160-64). Carboxymethylation of these two residues did not affect the activity of the enzyme to any significant extent. However, when M15, a dimeric alpha-acceptor molecule, was treated with iodoacetate, three additional cysteine residues at positions 76, 389 and 602 were modified. Carboxymethylation was found to inhibit the ability of M15 to participate in alpha-complementation. This suggests that one or more of these additional residues is situated at the dimer-dimer interface, the modification of which interferes with alpha-complementation.
In the present invention, it has been discovered, quite unexpectedly, that modification of certain of the amino acid residues resulted in a significant increase in activity of the enzyme. This increase in activity occurred not only with respect to the complementation activity, which is the rate at which the catalytically-inactive donor and acceptor dimers combine to form the catalytically-active .beta.-galactosidase tetramer, but also with respect to the equilibrium activity, or the rate at which the reformed .beta.-galactosidase tetramer converts a substrate. By increasing .beta.-galactosidase activity, the sensitivity of an assay for an analyte is increased, and such increase in sensitivity means that certain analytes which are present in blood serum in very minute quantities, such as drugs and drug metabolites, can be detected more accurately. Using the muteins of the present invention, the sensitivity of an assay can be increased as much as 1.3- to 3-fold or greater without a significant loss of enzyme stability.
Predetermined, site-directed mutagenesis of tRNA synthetase in which a cysteine residue is converted to serine has been reported (G. Winter et al., 1982, Nature, 299, 756-758, and A. Wilkinson et al., 1984, Nature, 307, 187-188). Estell et al., U.S. Pat. No. 4,760,025, issued Jul. 26, 1988 describe a cloned subtilisin gene modified at specific sites to cause amino acid substitutions of certain methionine residues. Koths et al., U.S. Pat. No. 4,752,585 issued Jun. 21, 1988 and U.S. Pat. No. 5,116,943, issued May 26, 1992, describe the protection of a therapeutic protein such as interleukin-2 or interferon-.beta. against oxidation by substituting a conservative amino acid for each methionyl residue susceptible to chloramine T or peroxide oxidation.
Buchwalter et al., European Appl. No. 91106224.8, published Nov. 27, 1991, describe an animal somatotropin derivative in which cysteine has been substituted by site-specific mutagenesis techniques for certain serine and tyrosine residues and in which glutamic acid has been substituted for certain cysteine residues. Breddam et al., PCT/DK91/00103 published Oct. 31, 1991, describe chemically modified detergent enzymes wherein one or more methionines have been mutated into cysteines, and then said cysteines are subsequently chemically modified in order to improve stability of the enzyme toward oxidative agents. Mattes et el., U.S. Pat. No. 4,963,469, issued Oct. 16, 1990, describe a change of an amino acid in the region between amino acid 430 and 550 of .beta.-galactosidase to another amino acid to produce an enzymatically inactive, immunologically active .beta.-galactosidase mutein. Estell et al. (1985, J. Biol. Chem. 260, 6518-6521) used site-directed mutagenesis to alter the methionine 222 residue of subtilisin which is a primary site for oxidative inactivation of the enzyme. These authors found that mutants containing non-oxidizable amino acids, i.e., serine, alanine and leucine, were resistant to peroxide inactivation, whereas methionine and cysteine-substituted enzymes were rapidly inactivated.
In copending and commonly assigned application U.S. Ser. No. 08/146,633 filed Oct. 29, 1993, the content of which is herein incorporated by reference, it was discovered that substitution by site-directed mutagenesis of the cysteine-602 residue on an enzyme acceptor polypeptide fragment of .beta.-galactosidase with a conservative amino acid, preferably serine, results in substantially increased stability of the enzyme acceptor mutein over that of an enzyme acceptor polypeptide fragment having cysteine at position 602.
As used herein, the numbering for the amino acid residues of .beta.-galactosidase will be that published by Kalnins et al., 1983, EMBO Journal 2, 593-597, the content of which is herein incorporated by reference. The nucleotide sequence of the lac Z gene coding for .beta.-galactosidase in E. coli was determined and .beta.-galactosidase was predicted to consist of 1023 amino acid residues rather than the 1021 residues previously reported by Fowler and Zabin (1977, Proc. Natl. Acad. Sci. USA 74, 1507-1510 and 1978, J. Biol. Chem. 253, 5521-5525).